The structure of your data team, and why it matters

How data teams fit into organizations

How does an organization make sure that the projects that it undertakes are successful? If we were to trace the problem noted above to its roots, we would find, among other things, a lack of collaboration. Between leaders and employees, between teams, and between team members.

This lack is often a structural issue. Data scientists, data engineers and analysts exist in silos, and no-one is owning the results. If no-one is owning the results, no-one is facilitating communication, tracking progress, and making sure that it is relevant to the problems that the organization is trying to solve.

When planning to integrate a data science team, companies face the decision of either a centralized, or decentralized strategy (while other models exist, these are the most straightforward).

Decentralized

A decentralized approach is often the least coordinated, typically because the data science function emerged based on an existing need within the business.

The drawbacks to this model being:

• It can lead to silos, decentralized reporting, and a lack of standardization.

• Lower quality standards are likely. Data scientists may be left on their own, without gaining knowledge from mentoring scientists.

• Harder to build individual development plan for career growth of scientists working independently.

Centralized 

In the centralized model, a separate, core data science team is in place, with resources allocated based on the needs of individual business units. This core team serves the organization as a whole on a variety of projects. The advantages being that it is easier to track career growth and allows for better resource management.

Common challenges faced in this model:

• A disconnect between the core data team and individual business units. The former may not be aware of the latters needs and pain points unless there is proactive communication.

• Conflicts in scheduling. A solution proposed by the core team may not fit into a business units roadmap, leading to frustration.

What we would suggest:

For a smaller organization just starting out with building a data team, the centralized/hybrid model is easier to implement, with less moving parts. Individual or smaller groups of data scientists can then be embedded into different business units, so that they are working on projects that are valuable to the business. This may lead to knowledge silos, so it is important to make sure that knowledge is periodically shared between these embedded scientists.

🎙️ A chat with Alan Roan of Cervus.ai pt.2

In the last issue, the Managing Director of Cervus.ai, Alan Roan, answered some questions on their recruitment process and the experience they’ve had with graduates as a startup. Here’s the 2nd part of that interview:

When does experience exceed talent (or visa versa) when looking at a new hire?

More importantly is fit. I always ask, will this person fit into our culture, which trumps both talent and experience. The sad thing is that the traditional HR process relies on the submission of a CV that scores an individual on talent and experience alone.

How do you know whether someone has the skills required to work effectively within a team? What is the most important factor when listing requirements?

You don’t, you nurture and train them so they can gain the correct skills and work effectively. However, it is far more important to have the ability to work hard and fit in with others.

What is it graduates struggle with most when joining Cervus.ai?

Graduates struggle most with working in an unfamiliar environment.

📊 Future Focused

Organizations are using tech to fast track their hiring process.

AI is being used to automate many aspects of the talent acquisition process, from resume screening to candidate matching. AI can augment recruiters' abilities and free up their time to focus on more strategic activities. Organizations are using Applicant Tracking Systems (ATSs) to streamline the hiring process and make it more efficient. These systems allow recruiters and hiring managers to easily track candidate applications, schedule interviews, and manage the hiring process from start to finish.

Video interviewing, virtual career fairs and social recruiting are becoming increasingly popular as a way to assess candidates remotely. By leveraging video conferencing technology, recruiters and hiring managers can conduct interviews with candidates from anywhere in the world, saving time and resources.

Social media also helps organizations to build their brand, connect with the talent pool, and advertise job openings to a wider audience.

Many companies report continued challenges in hiring, risk management, and developing AI / ML skills in their tech and non-tech staff.